A New Argument for Lexical Decomposition: Transparent Readings of Verbs

Percus (2000) has argued that subordinate verbs do not admit of transparent (de re) readings. In this squib, I will argue that such verbs in English do admit of a kind of transparent reading. Furthermore, this particular reading suggests that, at some level of syntactic and/or semantic analysis, verbs must be decomposed into a primitive action predicate (i.e., do) and a nominal argument describing the action performed. That is, this reading suggests that a verb like juggle must be semantically decomposed as a complex predicate do juggling.

To keep my discussion here brief, I will presuppose some familiarity with Percus 2000, as well as with the kind of situation-semantics framework Percus assumes (see Kratzer 2008). In the following section, I introduce the reading of interest.

1 A Transparent Reading of Subordinate Verbs

Imagine the following scenario. You and I are roommates. Moreover, we have entered ourselves into a talent competition. The act that we are preparing to perform is, specifically, a juggling routine. Every day at 3 p.m., we go to the gym to practice our juggling.

Enter our friend Mary. Mary lives with us and knows that we are in a talent competition. However, Mary doesn't have any idea what kind of act we will be performing. That is, although Mary knows that we leave for the gym every day at 3 p.m. to practice our act, she has no idea that we are, specifically, juggling.

Now, suppose that one day, we decide to blow off practice. That is, at 3 p.m., we tell Mary that we're leaving to practice, but then decide at the last minute to go to the movies. Suppose that, at the theater, I turn to you and utter sentence (1a).

1. 1.

   1. a. Mary thinks we're juggling right now.

   2. b. Mary thinks we're doing (our daily) juggling right now.

My own judgment, as well as the judgment of other English speakers I've polled, is that there is an interpretation under which sentence (1a) is true in the scenario described. Furthermore, let us also briefly note that sentence (1b) likewise possesses an interpretation under which it is true in this scenario. Finally, let us note that, for what it's worth, speakers accept (1b) as a paraphrase of the interpretation under which (1a) is true.

In a moment, we will begin to consider what kind of logical formula could represent this interpretation of (1a). Before we do, however, let us consider a somewhat more complex case, as it will prove helpful to the development of the analysis.

Suppose now that my brother has joined our little juggling troupe. Suppose also that Mary knows this, and knows that he practices with us every day at 3 p.m. Again, though, Mary has no idea what kind of act we are performing. Finally, suppose that one day, my brother blows off practice. That is, he arrives at our apartment, tells Mary that we're all going off to practice, but then at the last moment goes to the movies instead. Now suppose that at the gym, you mistakenly remark that Mary might be mad at my brother for going to the movies without her. As a correction, I utter the sentence in (2a).

1. 2.

   1. a. Mary thinks my brother is juggling right now.

   2. b. Mary thinks my brother is doing (his daily) juggling right now.

Again, the judgment from English speakers seems to be that (2a) is both natural and true in the scenario described. Furthermore, this is again paralleled by the sentence in (2b), which is also true in this scenario, and which speakers accept as a kind of paraphrase of their interpretation of (2a).

2 Characterizing the Reading

Let us, then, consider what kind of logical formula could represent this reading of (2a).¹ First of all, we can rule out the formula in (3).²

1. 3. think(s₀, Mary, [λs. juggle(s, my-brother(s))])

This formula represents a simple, purely opaque reading of sentence (2a). According to this formula, in all situations s consistent...